On Sun, Apr 12, 2015 at 1:00 PM, Ted Dunning <ted.dunn...@gmail.com> wrote: > Roman, > > Could you comment on the potential issue that this proposal is only > offering up the distributed object storage. The problem that might arise > with this is that by offering up a particular specialized core capability, > other projects may find it difficult to use this core capability which > would inhibit growth of the community. > > What do you think about that?
I was trying to address that very point in the Rationale section. While it is true that today Geode is the engine for GemFire, it is at the same time a pretty sophisticated piece of engineering work with subsystems that can find use outside of GemFire applications. Just to give you a few examples: there's a kick-ass eventing systems in there and the off-heap memory management implementation is one the best ones I've seen. I could absolutely see Geode integrating much tighter at the level of in-memory data exchange layer (think Tachyon done right) with the rest of the bigdata ecosystem as well. That said, making Geode as modular as possible is exactly why we're proposing it for Incubation: this is exactly the kind of extended use cases that ASF communities are great at figuring out. Long story short, it is definitely our desire to see Geode applicable way outside of core GemFire capabilities and I believe it has got what it takes. Thanks, Roman. P.S. In fact, you've just given me a great subject for a blog post. Stay tuned! --------------------------------------------------------------------- To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
